Quote:

also other OS have since long time memory protect, but no OS4 library interface, so its sure this is not important to use.

OS4's library interfaces have nothing to do with memory protection.

They're a good idea, though, since they simplify things like plugin interfaces, local library instances in threaded code, and eliminate namespace clashes. This is why I use them, and think they're useful.

The fact that OS4 libraries use interfaces by design means that I can easily have per-thread interface instances when and where necessary. I find this very useful in PFTP, for example.


Quote:

but when look on available OS4 sources there is this additonal typing in.

Oh no! How will we ever cope? If only it was possible to copy things you've already typed in and paste them elsewhere. Or to search for strings and replace them.


Quote:

and this is what other devs see.the OS need lots of unnecessary typing and think wy i should choose this OS

I seriously don't think that OS4 has any less developers because of library interfaces.


Quote:

So i think, most think Hyperion have done wrong all can do, and so there are not many that buy or code for OS4 and this result in Software situation and few software.

It appears to me that OS4 has one of the liveliest developer communities of the Amiga world right now.


Quote:

in the last 3 years not much is add to OS4.

In the last 3 years, I have received beta updates pretty much every single day. OS4's development is not stagnated at all.

Quote:

also what OS4 miss, parameter checking,(AFA have this on some function add in safer system mode) this is too important for a memprotect OS, is easy to implement and avoid lots of crashes.

OS4 has a lot of parameter checking. More than OS3.x did. Of course, I've seen the changelogs, so I actually know what i'm talking about. But believe what you want.

_________________
Playstation Network ID: xeron6

@xeron

>OS4 has a lot of parameter checking. More than OS3.x did. Of course, I've seen the >changelogs, so I actually know what i'm talking about. But believe what you want.

oh the mystic better OS4, as far i know OS4 bring reaper when do a FreeMem with address 0 or a DeletePool with pool address 0.so there is no parametercheck and handle it without program crash.

but on AOS a deletePool with poolheader 0 address or freemem with 0 address do nothing(not crash the program) because the wrong 0 parameter is handle by the OS correct

@bernd_afa

Wow. You know specifically two functions which don't have parameter checking in the current public release of OS4, therefore the OS4 devs must be completely unaware of parameter checking in general.

Besides, if you write your software according to the autodocs, you won't pass zero to those functions, and you won't get a reaper.

_________________
Playstation Network ID: xeron6

@bernd_afa

I think it's pretty clear that you don't understand what interfaces on OS4 are about. The idea is that instead of having a single jumptable for a library, as is the case with the classic Amiga shared library system, you can have several totally independent interfaces of the same library. This means that in the future completely clean and modern interfaces can be added to libraries, and new software can use these interfaces, while old software can keep using the old interfaces. It's also possible to protect critical functions, so that they cannot be patched, which is an important safety measure (with the classic Amiga shared library system any function of any library could be patched which makes it extremely easy for a virus to do damage to the system). So the OS4 library interfaces system provides a lot more flexibility and it will allow clean new APIs with memory protection to be added in the future, while retaining the old interfaces for compatibility.

@xeron

It´s useless, bernd has his agenda and he fight for it with all methods, including false information and hearsay.

I bet he also didn´t know that the handling of strictly forbidden double use of IORequests with e.g. timer.device is done the same way in OS3 and OS4. I also bet that he don´t know it´s strictly forbidden, but OS4 devs were kind enough to implement the workaround for such foul programs again...

Regards

@bernd_afa

Quote:

oh the mystic better OS4, as far i know OS4 bring reaper when do a FreeMem with address 0 or a DeletePool with pool address 0

Of course it does, trying to free memory that has not been allocated is a software fault, and when some software does illegal things, it is the job of the OS to catch it, and report to the user that a program is misbehaving. What you have described is the correct behaviour.

@whose

Additionally, OS4 provides an officially sanctioned way to duplicate IORequests as well

_________________
Playstation Network ID: xeron6

@xeron

it provides some more very interesting functions, as I have seen

Regards

Star Wars Darth Vader theme music playing.....deep breathing

I have foreseen a most ambitious project : The Amiga Inc. rebels propose an end to the lawsuit with Hyperion. AI proposes making a “Amiga” look-alike for Google’s Android platform aka Amiga OS5.0. . Hyperion gets receives a very large advance check and gets to keep all rights to Amiga OS4.0. Rogue has been seen spending sleepless nights after reviewing the SDK

Last edited by Interesting on 15-Apr-2009 at 03:52 PM.

_________________
"The system no longer works " -- Young Anakin Skywalker

@cobra

>It's also possible to protect critical functions, so that they cannot be patched, >which is an important safety measure (with the classic Amiga shared library >system any function of any library could be patched which makes it extremely easy >for a virus to do damage to the system).

nothing is easier than that on 68k system too.use mmu.library, protect the adress space of the jump tables with MMU. so programs that want patch are stop with illegal address access.

this can also done on every UAE system that have a memory protect Kernel.(Linux and windows)by adding the address space of the jump tables to a hidden area or the code of the library.no extra interface need.

having diffrent library versions is a good idea, is this only a announce, or does it really work ?.
Does this work on OS4 to have for example a mpega.library V43 and another program running mpeg.library V21 ?

but it is also no problem to change 68k openlibrary to support this.

all libraries are in a linked list.current open library code doesnt care about lib version.

but when change lib open code, so it also check for lib version and if lib version is lower than current it can then load the lib and give the program an additional pointer.

>Of course it does, trying to free memory that has not been allocated is a software >fault, and when some software does illegal things, it is the job of the OS to catch it, >and report to the user that a program is misbehaving. What you have described is >the correct behaviour.

That is a step backward in computing.see linux or windows programs or 3.1 programs.when you stop a program that have a 0 adress for memory then near no windows or linux program work.try it öut, remove in unix free command this line.

free (void *mem)
{
struct mem_block *block;
u_int *end_magic;
int omask;
usetup;

>>>> if (! mem) return;

sure you can say this feature is in RKM only for freevec in.but what should OS4 be. a strip down AOS that support only features that are in the rkm written 1990 ?

AOS 3.x support it and many programs crash when you not support it.

it is not necessary to write always

if (mem)FreeMem(xx);

if you realy want new coder bring to OS4, then this is the best way that they think.are they crazy, every OS support 0 pointer check and do nothing so a if (mem) can avoid.

and for such a OS with that features they want so much money ?

Last edited by bernd_afa on 15-Apr-2009 at 05:04 PM.

@bernd_afa

Quote:

nothing is easier than that on 68k system too.use mmu.library, protect the adress space of the jump tables with MMU. so programs that want patch are stop with illegal address access.

Strange this wasn't already done then if it's so easy...

_________________
AOS 4.1 : I dream it, Hyperion did it !
Now dreaming AOS 4.2...
Thank you to all devs involved for this great job !

@bernd_afa
Quote:

oh the mystic better OS4, as far i know OS4 bring reaper when do a FreeMem with address 0 or a DeletePool with pool address 0.so there is no parametercheck and handle it without program crash.

No crash here on my 4.1 install. And yes, AmigaOS 4.1 is better than previous versions and it will become even better as development has never stopped. Live in the past all you want but don't drag us all back into the dark cave with you.

_________________
ExecSG Team Lead

@abalaban

>Strange this wasn't already done then if it's so easy...

i like AOS because it grow with the patches.I am not limit to a old fashion workbench, i use dopus magellan and 32bit support, but there is also Scalos possible.

when cant patch, then much functionality is use.i also use a macro recorder machv.here i can hotkeys and toolbar buttons for every program make and many other patches that make using of AOS more comfortable.

so avoid patching of OS functions is no good idea.also windows and linux allow patches.

amiga OS and amiga OS4 is not safe, a memprotectet library is realy no barrier for viruses.

@ssolie

>Live in the past all you want but don't drag us all back into the dark cave with you.

and thats what i called the OS4 fanatic.

User that tell that all that have no OS4 live on the past.68k amiga OS is furtherdevelop continusly MOS and AROS is furtherdevelop.only because there is native 68k coldfire V5 out witzh 500 MHZ and nobody build a 68k with 1 GHZ is really no reason to port my software from 1 retro OS to another retro OS.

I think All amiga OS are retro OS, they are only attractive for very few users.
I am not so blind and think any amiga OS have a chance ti get attractive if this OS are so slow develop.

and this include MOS and OS4 too.they are commercial, but dev speed is slow too.
and for my matter of taste toooooo slow that i buy it.

its realy sad that there is no working together possible, because there are devs that want fight for thier OS and want that all should buy this OS.

Last edited by bernd_afa on 15-Apr-2009 at 05:22 PM.
Last edited by bernd_afa on 15-Apr-2009 at 05:20 PM.
Last edited by bernd_afa on 15-Apr-2009 at 05:19 PM.

@bernd_afa

Quote:

also windows and linux allow patches.

To my knowledge Windows allows to install some sort of "callbacks" (to simplify) and for Linux I don't know how except by replacing the library itself you would be able to patch their functions (at runtime per se, else I'm fully aware of linux source code patchs but that's another thing than what we are discussing here).

For the remaining of your post I'm not sure I understood but for me DopusMagellan (as Scalos) is not a patch, but a replacement which is different.

Take for example an SSL library you don't want someone to be able to patch it behind you breaking your encryption do you ? AOS4 allows a library (so it's on a case by case basis) to specify it does not want to be patchable.

_________________
AOS 4.1 : I dream it, Hyperion did it !
Now dreaming AOS 4.2...
Thank you to all devs involved for this great job !

@ChrisH & all

I've got another idea. Hyperion are going to sell AmigaOS4 as an embedded solution. With ACube we already have embedded based hardware. Hyperion have stated before that OS4 was not targetted to be a desktop OS, excepting us that wish to use it that way. So perhaps they are going to go public with it? In this case, your nor competing against Mac or Windows anymore (not that we were). but against Linux. Which has a strong hold in a lot of those areas.

@bernd_afa
And thats what i called the OS3 fanatic. Now can you please stop your anti-OS4 diatribe, and do something constructive? Or at least please don't do it IN MY THREAD, thanks in advance (hint: you only get one polite warning).

Last edited by ChrisH on 15-Apr-2009 at 05:57 PM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@ChrisH

>And thats what i called the OS3 fanatic. Now can you please stop your anti-OS4 >diatribe, and do something constructive? Or at least please don't do it IN MY >THREAD, thanks in advance (hint: you only get one polite warning).

yeah you are really great you ask what people think in the beginning thread.if a new memprotect system is really come some day from Hyperion.then i write what i think and then you say to me i do anti OS4 debatte.

sorry that i cant write more positive about the Hyperion activity, but when you open your eyes and see how many currently use OS4 or code for it, from the Million market of amiga that want AOS go to new Hardware, because real new 68k CPU are not availible, then you maybe see that i only write realistic.maybe it is understand wrong because of my bad english, i can explain more, but all i want do is constructive.

Hyperion should announce only things that are realistic reach.i still remember that HJF tell in a interview that there are intrestings of using OS4 in Kiosk systems.

Now 7 Months later, seem nobody want OS4 for this because, i read no message about that.

this say too, OS4 is not good enough so i be constructive and tell what they should do better.

Hyperion should port their OS to a cheap and powerful Hardware as other OS have, because what reason is there to buy for more money a OS4 Hardware and OS4 for more money if no OS4 Fan ?.And many OS4 Fans are not here, i dont think that more than 1500 have OS4.

luckily OS4 is not the only OS in the world.but bad for OS4, because it must compare in Feature and price with other OS.

Last edited by bernd_afa on 15-Apr-2009 at 07:34 PM.
Last edited by bernd_afa on 15-Apr-2009 at 07:32 PM.

@ChrisH:

Just report him... debate is futile, he won´t stop with writing off topic.

I think Bernd makes some good points, particularly re: source compatibility as it relates to the community's insistence on downlevel application support in OS4. Bernd is saying, "Hey, I want to port my apps, but I don't want to maintain a separate source tree for OS4 when I can compile as-is for OS3, MorphOS, and AROS." Or something like that.

Namespace conflicts aren't a good enough argument for decorated function calls. The others--encapsulation, polymorphism, etc.--are good arguments. I don't think the current API is object oriented enough for it to make a difference, though. Someone with more OS4 knowledge than me needs to chime in here.

I think Bernd is wrong re: null pointers. Raising an exception is a Good Thing, as passing a null pointer to an OS function is indicative of a flaw in the application. Laziness, i.e. not wanting to validate data or catch and process exceptions, is not a good excuse, and the OS itself should be kept as simple and concise as possible. (At least, that's one train of thought. Java, .NET, et al have made programmers a bit greedy. They want garbage collection, thought-free exception handling, and other features brought down to the OS level. There's both good and bad in that.)

EDIT: Bernd's English is bad, yes, and I don't agree with all of his ideas, but he's only expressing his frustration with the current state of things. Why does everyone try so hard to silence the opposition? Debate is healthy. Do we have an issue with cultural differences? Something lost in translation (literally)?

Last edited by Trev on 15-Apr-2009 at 07:56 PM.
Last edited by Trev on 15-Apr-2009 at 07:54 PM.

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

@bernd_afa

Quote:

IExec->AllocMem is pure C code.when you look at bars&Pipes plugin interface they do the same calling conventions.but because other OS do not use it, and it need extra letters and need extra macros to make it easy so other OS dont use it.

Yes, this true... and not true...

I used function pointers inside structures in PublicPaint, the idea was create a set of optimized function for 32bit, 16bit and 8bit.

I used this idea long before OS4 adapted the idea.

Quote:

GFX = FUNCTIONS_32BIT; GFX->SETCOLOR(255,255,0); for (x=0;x<100;x++) { for(y=0;y<100;y++) { GFX->PLOT(x,y); } }

the only change needed to draw the same in 16bit, is the first line, whit simple inline macros you hide the GFX-> part, this is what AmigaOS4 does, but some developers like the GFX-> part and use it, that's there personal choice, you are not forced write it one way or the other.

No this has nothing to do whit memory protection, but enables some flexibility over the static JMP table, what it does is enable reversion whit out taking code out, and breaking legacy.

If you look at Linux /lib directory you see lots of different versions of same .so files, there are always a problem when you have programs whit different version requirements, hopefully interfaces can solve this problem for OS4, well theoretical at least.

Last edited by NutsAboutAmiga on 15-Apr-2009 at 08:03 PM.
Last edited by NutsAboutAmiga on 15-Apr-2009 at 08:02 PM.
Last edited by NutsAboutAmiga on 15-Apr-2009 at 07:54 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS